Background: Both the professional and the general media have recently published concerns about an "epidemic" of child and adolescent depression. Reasons for this concern include (1) increases in antidepressant prescriptions, (2) retrospective recall by successive birth cohorts of adults, (3) rising adolescent suicide rates until 1990, and (4) evidence of an increase in emotional problems across three cohorts of British adolescents. Methods: Epidemiologic studies of children born between 1965 and 1996 were reviewed and a meta-analysis conducted of all studies that used structured diagnostic interviews to make formal diagnoses of depression on representative population samples of participants up to age 18. The effect of year of birth on prevalence was estimated, controlling for age, sex, sample size, taxonomy (e.g., DSM vs. ICD), measurement instrument, and time-frame of the interview (current, 3 months, 6 months, 12 months). Results: Twenty-six studies were identified, generating close to 60,000 observations on children born between 1965 and 1996 who had received at least one structured psychiatric interview capable of making a formal diagnosis of depression. Rates of depression showed no effect of year of birth. There was little effect of taxonomy, measurement instrument, or time-frame of interview. The overall prevalence estimates were: under 13, 2.8% (standard error (SE) 0.5%); 13-18 5.6% (SE 0.3%); 13-18 girls: 5.9% (SE 0.3%); 13-18 boys: 4.6% (SE 0.3%). Conclusions: When concurrent assessment rather than retrospective recall is used, there is no evidence for an increased prevalence of child or adolescent depression over the past 30 years. Public perception of an "epidemic" may arise from heightened awareness of a disorder that was long under-diagnosed by clinicians.